A Pragmatic Intervention to Increase Screening of High-Risk Individuals for Diabetic Retinopathy

Abstract

Diabetic retinopathy (DR) is a leading cause of blindness among U.S. working-age adults. Hispanic and Black individuals with diabetes are disproportionately impacted. Early detection of DR can prevent vision loss; however, at least 30% of patients at risk are not screened. Data were collected in a primary care clinic from September 2023 – August 2024. People with diabetes overdue for an eye exam were referred for photo screening during a primary care visit. Referrals, screening completion, interpretation of results, and follow-up eye visits were recorded in the electronic health record (EHR). After screening, individuals were asked to complete an experience survey. Of the 124 individuals referred during their primary care visit, 78% (n=97) completed the screening. Of those screened, 77% (n=75) resulted in interpretable findings. Of these 75 screened patients, 9% (n=7) detected severe (vision threatening) DR and 24% (n=18) detected mild to moderate DR. 24% completed the experience survey. 22 respondents (96%) reported that it was very easy to get their eyes screened. Half reported that they would not have scheduled a separate screening visit if it hadn’t been offered as part of their visit. All respondents recommend the intervention. We detected significant eye findings, including vision threatening retinopathy among a population likely to not have been otherwise screened. Patients reported positive experiences and recommended this approach. That we accomplished this at a primary care clinic location with a higher percent of patients of color than our general patient population indicates a contribution to equitable diabetes outcomes.

Keywords

diabetic retinopathy / prevention & control eye screening health equity primary health care / methods retina / diagnostic imaging health promotion

Background

Diabetes is a leading cause of blindness among American adults due to the damage it can cause to the retina, known as diabetic retinopathy (DR). Hispanic and Black patients with diabetes are 2-3 times more likely to develop severe DR than their white peers.^1-3

Early detection of DR is key to preventing irreversible vision loss. The gold standard for retinopathy screening is an annual dilated eye exam with an ophthalmologist or optometrist.⁴ When DR is identified, interventions exist to preserve vision. In mild cases, monitoring and management of diabetes and blood pressure can prevent progression. In more severe cases, medication, laser treatment, or surgery may be needed.^5,6

Nationally, just 69% of patients at risk of DR are screened.^7-9 Patients with diabetes who do not receive regular eye exams are more likely to have sight-threatening retinopathy as the condition has progressed and impacted vision. People of color are less likely than white patients to get an annual eye exam due to known barriers including time off work, childcare, transportation, misunderstanding of insurance coverage and healthcare cost concerns, hesitancy to have pupils dilated, low health care literacy, lack of racial or ethnic concordance with providers, and lack of understanding that early diabetic damage to the retina is asymptomatic yet risk of vision loss can be minimized.¹⁰

Primary care is more equitably accessed by patients and has fewer barriers to access compared to specialty care and therefore is an ideal place to intervene to address health disparities.^11,12 Primary care visits have time constraints, however, so favor quick and impactful strategies.

Purpose

An ophthalmologist in our integrated health system identified an opportunity to address a health disparity via primary care intervention. By forming a cross-organizational partnership and securing pilot funding from a hospital foundation to purchase new technology (a non-mydriatic fundus camera) for taking photos of the eye during a primary care visit, the project team was able to implement and evaluate a novel approach to DR screening.

In a clinic that serves over 11,000 patients annually, including over half identifying as people of color, nearly one-third diagnosed with diabetes, and an opportunity to improve DR screening rates, we explored the feasibility and potential impact of using eye photographs taken during primary care visits to screen for DR. Specifically, patient experience and early indicators of this intervention’s impact, overall and by patient characteristics, are reported.

Methods

Setting

This pilot was conducted within a Midwestern integrated health system offering care, coverage, research and education. One inner-ring suburban primary care clinic with co-located eye care was selected as our pilot site because the patient population has a higher proportion of people of color, and more uninsured or underinsured patients than our other locations. In addition, the Healthcare Effectiveness Data and Information Set (HEDIS)¹³ eye exam measure had very low adherence among the ∼2,500 individuals with diabetes empaneled at the clinic. Importantly, the clinic’s primary care leadership team was committed to partnering with ophthalmology to meet this opportunity. Finally, optometric technicians work on site and were able to be trained in the new eye photo technology that the American Academy of Ophthalmology deemed an effective alternative to a dilated eye exam in 2024.⁴ This project was determined to not meet the definition of human subjects research under 45 CFR Part 4 in May 2023 and thus be exempt from IRB oversight by the HealthPartners Institute Institutional Review Board.

We describe photo screenings completed at this pilot site between September 2023 – August 2024.

Implementation

In early 2023, the project team worked with IT programmers to create an EPIC workflow that allowed primary care providers to refer a patient to the eye clinic for a photo immediately following their visit. The workflow also prompted the optometry technician’s order and created a place to upload the image.

Before the intervention, any patient with diabetes who had a primary care visit and was flagged in EHR as not having had an eye exam in the last two years received a verbal recommendation from their provider to schedule and complete an eye screening. In September 2023, this intervention to refer these patients to complete a photo screening on site following the primary care visit was introduced and continued until November 2024. All primary care providers at the pilot site learned how to use the new order before the pilot launched. During this pilot, neither patients nor insurance were billed for this intervention.

The photo screening was considered to be a routine part of eye care in which verbal consent is obtained as part of standard care. As such, the HealthPartners Institute Institutional Review Board granted a waiver of documentation of informed consent for the photos and survey completion as the project presents no more than minimal risk of harm to participants and involves no procedures for which written consent is normally required outside of the research context (HealthPartners Institute Institutional Review Board; Approval Number-NHSR 23 05 011). If the patient verbally agreed to a photo screening, they were scheduled to be seen at the onsite eye clinic immediately following their primary care visit. The technicians onsite were trained to use the camera (an eye non-mydriatic fundus camera – Nidek AFC 330), upload the image to EPIC, and create an inbox task for the reading ophthalmologist.

The ophthalmologist (JAB) interpreted those photos, reported findings in patient charts with a copy to primary care clinicians, and messaged and/or called patients with results. The messages recommended appropriate next steps based on findings: to either schedule a complete dilated eye exam soon after notification, or to follow-up with optometry or ophthalmology in one year.

In November 2023, project evaluators suggested developing EPIC SmartText to standardize patient messages from the ophthalmologist. This brought efficiencies and improved the ability to monitor intervention fidelity over time and summarize at the end of the pilot.

Electronic Health Record Data – Outcomes

To understand intervention fidelity, screening referrals and outcomes documented in EPIC included: primary care referral, completion of phone screening with eye technician, results of conclusive screenings (level of DR damage detected, other eye concerns detected, or no concerns detected), inconclusive screenings, and ophthalmologist recommendation: to schedule a complete dilated eye exam or to follow-up in one year. We also summarized attendance at follow-up appointments through November 2024. Eye appointment receipt was defined as a documented eye appointment with departments whose name contained ophthalmology, optometry, eye, lens or retina.

Claims Data – Outcomes

For patients who are also members of one of our health plans, we looked at claims data for CPT codes indicating an eye exam was completed but not in our EHR, likely because it was conducted outside of our health care system. We did not limit the time period for completing a follow up appointment; data were pulled in November 2024.

Patient Survey Development

We also developed a patient experience survey to inform process improvement and assess intervention feasibility (see Supplemental Material). The survey was guided by the Theoretical Framework for Acceptability, and included items to assess affective attitude, burden, intervention coherence, perceived effectiveness, and self-efficacy.¹⁴ The survey also included items about knowledge of health consequences, informed by COM-B theory.¹⁵ The survey, which included both open- and close-ended questions, was created in partnership with study team members using best practices in survey design.¹⁶ It was reviewed for face validity by study team members inclusive of a survey methodologist, ophthalmologist, and diabetes specialist not involved in its initial development. Given the linguistic diversity of this patient population, the 10-item survey was also adapted to a shorter 7-question version to be completed by the patient with interpreter support.

Survey Data Collection

Originally, the survey was only offered online using a QR code patients could use to complete the survey on their phones. Uptake was minimal, so in December 2023, the team created a paper version and asked eye technicians to encourage survey completion onsite which increased survey response. If a patient did not have time to complete the survey onsite, they were given the QR code to complete later. Survey participation was voluntary and offered after the eye screening was completed. No incentives were offered to patients for survey completion. Data collection closed August 2024.

Analytic Approach

Descriptive statistics to summarize patient outcomes using EHR and claims data, as well as patient perspectives via the survey, are provided. Summary statistics describe available case analysis at the individual item level based on the completed responses only. Missing data were not imputed. Analysis was conducted using SAS v9.4.

Results

Population

Of the 124 individuals referred to complete an eye photo over the intervention period, average age is 59 years; 64% identify as male; 83% were not flagged as needing interpreter support; and 97% have type 2 diabetes. The majority identify as people of color, including 33% Black/African American, 17% Asian, and 14% Hispanic/Latino (see Table 1).

Table 1.

Patient Characteristics

	N=124 n (%) or mean (sd)
Average age	58.9 (13.4)
Gender
Male	79 (64%)
Female	45 (36%)
Interpreter status
Need interpreter support	21 (17%)
Do not need interpreter support	103 (83%)
Diabetes type
Type 1	1 (1%)
Type 2	120 (97%)
Unknown	3 (2%)
Race/ethnicity
Asian	21 (17%)
Black/African American	41 (33%)
Hispanic/Latino	17 (14%)
White	43 (35%)
Other/did not respond	2 (1%)

Screening Outcomes

Of the 124 patients referred during their primary care visit (Table 2), 78% (n=97) completed the screening. Of those screened, 77% (n=75) had interpretable findings for DR or other eye conditions. Uninterpretable results occurred because the camera did not capture a clear image of the retinal blood vessels. Some uninterpretable results for DR detected cataracts or other eye health concerns (n=11); these were reported to the patient with a recommendation to schedule a complete eye exam.

Table 2.

Intervention Results

	n	%
Referred to complete an eye photo	124	100%
Completed an eye photo	97	78% of referred
Received uninterpretable results	22	23% of completed photo
Received interpretable results	75	77% of completed photo
Severe (vision-threatening) damage due to DR detected	7	9% of interpretable results
Mild to moderate damage due to DR detected	18	24% of interpretable results
Cataracts or other eye health concerns detected	11	15% of interpretable results
No damage detected	39	52% of interpretable results
Received recommendation for follow-up complete eye exam	50	52% of completed photo
Did not attend follow-up	18	36% of received follow-up rec.
Attended follow-up due to:	32	64% of received follow-up rec.
Severe (vision-threatening) damage detected	5	16% of attended follow-up
Mild to moderate damage detected	9	28% of attended follow-up
Cataracts or other eye health concerns detected	6	19% of attended follow-up
Received uninterpretable results	12	38% of attended follow-up

Diabetic retinopathy was detected via the photo screening. Of the 75 interpretable photos, 9% (n=7) detected severe damage due to DR and 24% (n=18) detected mild to moderate damage.

While 58 individuals with detected concerns or uninterpretable photos should have received a recommendation to schedule a complete eye exam, only 50 did due to limitations in our processes early in the pilot. 32 of those who received the recommendation (64%) attended a complete eye exam appointment; 14 of whom were referred because their photo detected possible DR, 5 with potentially severe DR. In addition, 12 were referred due to inconclusive screenings and 6 due to evidence of other eye health concerns such as cataracts. Follow-up appointments were completed a median of 40 days after initial screening; only five individuals had a follow up appointment more than 90 days after the screening.

Of the 14 individuals referred for detection of DR in the photo screening, six identified as Black/African American, four as White, three as Asian, and one as Hispanic/Latino.

Of the 97 individuals who completed a screening and were invited to participate in the survey, 23 individuals did so (24% of screenings). Of those who completed the survey, 22 respondents (96%) reported it was very easy to get their eyes screened as part of their primary care visit. Half would not have scheduled a separate screening visit if it hadn’t been offered as part of their primary care visit. All would definitely or somewhat recommend getting an eye screening in this way. See supplemental material for more details on survey results.

Discussion

Our single site pilot has promising early findings. We detected significant eye findings, including vision threatening retinopathy among a population likely to not have been otherwise screened. Moreover, our rate of identifying both DR and severe DR was higher than expected in the population.¹⁷ Screening also identified other eye health concerns (e.g., cataracts) in this population. This approach has been shown to be highly cost-effective.¹⁸ Our results are from one site within one health system, limiting their generalizability. For the subset of patients who do not also have insurance through our health system, we are not able to report on any eye visits completed in other systems, impacting both those we could identify as eligible for the intervention and observations of outcomes, so follow-through is underestimated. Our study design did not include inquiry into eye exam outcomes for the 32 patients who attended a follow-up visit, nor reasons why 18 patients did not attend a follow-up visit when recommended, but both are opportunities for future work. While we prioritized clinical support before project launch, we screened fewer patients than anticipated. In future work, we plan to expand to a broader group of primary care clinics using a handheld fundus camera that is simple enough for rooming staff working in primary care to be able to take photos and upload them to EHR. We will also consider offering dilation to participants to increase successful screening. Calling patients with abnormal results and arranging specialty follow up will help severely affected patients get the treatment they need to prevent vision loss.

Our self-reported data is limited to a small convenience sample and resulted from data collection methods that were modified over time in this pilot study. As such, we cannot enumerate a response rate nor speak to the generalizability of responders to the population who experienced the intervention. That we were at the relative ceiling of positive experiences suggests that the intervention is acceptable to a reasonable cross-section of recipients. Future evaluation will consider more robust collection of experience outcomes.

Conclusions

Patients who completed a photo screening for diabetic retinopathy reported positive experiences and said they would recommend this approach to other patients with diabetes. The majority also followed through on recommendations for follow-up with high fidelity. That we accomplished this pilot at a clinic location with a higher percent of patients of color than our general patient population indicates a contribution to health equity in diabetes outcomes and represents a cost-effective strategy for early detection of DR.

Supplemental Material

Supplemental material - A Pragmatic Intervention to Increase Screening of High-Risk Individuals for Diabetic Retinopathy

Supplemental material for A Pragmatic Intervention to Increase Screening of High-Risk Individuals for Diabetic Retinopathy by Maren S G. Henderson, Rachael L. Rivard, Jennifer M. Dinh, Gregg D. Simonson, Jeanette Y. Ziegenfuss, and Jane A. Bailey in Journal of Primary Care & Community Health.

Footnotes

ORCID iDs

Maren S. G. Henderson

Jennifer M. Dinh

Jeanette Y. Ziegenfuss

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Park Nicollet Foundation.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

No copyrighted material, surveys, instruments, or tools were used in the research described in this article.*

Supplemental Material

Supplemental material for this article is available online.

References

U.S. Department of Health and Human Services (HHS) . Healthy People 2030. Increase the proportion of adults with diabetes who have a yearly eye exam — D-04. https://odphp.health.gov/healthypeople

Lundeen

Wittenborn

Benoit

Saaddine

. Disparities in receipt of eye exams among Medicare Part B fee-for-service beneficiaries with diabetes - United States, 2017. MMWR Morb Mortal Wkly Rep. 2019;68(45):1020-1023. doi:10.15585/mmwr.mm6845a3.

Liu

TYA

Rosettie

, et al. Prevalence and health care disparities of retinal conditions: a meta-analysis. JAMA Ophthalmol. 2026;144:378-387. doi:10.1001/jamaophthalmol.2026.0319.

American Diabetes Association Professional Practice Committee. 12 . Retinopathy, neuropathy, and foot Care: standards of care in diabetes-2025. Diabetes Care. 2025;48(Supplement_1):S252-S265. doi:10.2337/dc25-S012.

Hutton

Stein

Bressler

, et al. Cost-effectiveness of intravitreous ranibizumab compared with panretinal photocoagulation for proliferative diabetic retinopathy: secondary analysis from a Diabetic Retinopathy Clinical Research Network randomized clinical trial. JAMA Ophthalmol. 2017;135(6):576-584. doi:10.1001/jamaophthalmol.2017.0837.

Ross

Hutton

Stein

, et al. Cost-effectiveness of aflibercept, bevacizumab, and ranibizumab for diabetic macular edema treatment: analysis from the Diabetic Retinopathy Clinical Research Network Comparative Effectiveness Trial. JAMA Ophthalmol. 2016;134(8):888-896. doi:10.1001/jamaophthalmol.2016.1669.

Ravindranath

Bernstein

Fernandez

Ludwig

Wang

. Social Determinants of Health and Perceived Barriers to Care in Diabetic Retinopathy Screening. JAMA Ophthalmol. 2023;141(12):1161-1171. doi:10.1001/jamaophthalmol.2023.5287.

Fathy

Patel

Sternberg

Jr. Kohanim

. Disparities in adherence to screening guidelines for diabetic retinopathy in the United States: a comprehensive review and guide for future directions. Semin Ophthalmol. 2016;31(4):364-377. doi:10.3109/08820538.2016.1154170.

Maclennan

McGwin

Jr. Heckemeyer

, et al. Eye care use among a high-risk diabetic population seen in a public hospital's clinics. JAMA Ophthalmol. 2014;132(2):162-167. doi:10.1001/jamaophthalmol.2013.6046.

10.

Coney

Scott

. Racial disparities in the screening and treatment of diabetic retinopathy. J Natl Med Assoc. 2022;114(2):171-181. doi:10.1016/j.jnma.2021.12.011.

11.

Starfield

Shi

Macinko

. Contribution of primary care to health systems and health. Milbank Q. 2005;83(3):457-502. doi:10.1111/j.1468-0009.2005.00409.x.

12.

Schuldt

Jinnett

. Barriers accessing specialty care in the United States: a patient perspective. BMC Health Serv Res. 2024;24(1):1549. doi:10.1186/s12913-024-11921-0.

13.

National Committee for Quality Assurance [NCQA] . HEDIS Measures and Technical Resources. https://www.ncqa.org/hedis/measures/

14.

Sekhon

Cartwright

Francis

. Acceptability of healthcare interventions: an overview of reviews and development of a theoretical framework BMC Health Services Research. 2017;17(1). 10.1186/s12913-017-2031-8.

15.

Michie

van Stralen

West

. The behaviour change wheel: a new method for characterising and designing behaviour change interventions. Implement Sci. 2011;6:42. doi:10.1186/1748-5908-6-42.

16.

Fowler

. Survey Research Methods. Sage Publications; 2013.

17.

Zhang

Cotch

Ryskulova

, et al. Vision health disparities in the United States by race/ethnicity, education, and economic status: findings from two nationally representative surveys. Am J Ophthalmol. 2012;154(6 Suppl):S53-62e1. doi:10.1016/j.ajo.2011.08.045.

18.

Siegel

Ali

Zhou

, et al.

Cost-effectiveness of interventions to manage diabetes: has the evidence changed since 2008?

Diabetes Care. 2020;43(7):1557-1592. doi:10.2337/dci20-0017.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.25 MB

0.00 MB